EP3274689A1

EP3274689A1 - Method for analysing particles

Info

Publication number: EP3274689A1
Application number: EP16717422.6A
Authority: EP
Inventors: Cédric ALLIER; Pierre BLANDIN; Anais ALI CHERIF; Lionel Herve
Original assignee: Horiba ABX SAS; Commissariat a lEnergie Atomique et aux Energies Alternatives CEA
Current assignee: Horiba ABX SAS; Commissariat a lEnergie Atomique et aux Energies Alternatives CEA
Priority date: 2015-03-24
Filing date: 2016-03-23
Publication date: 2018-01-31
Anticipated expiration: 2036-03-23
Also published as: FR3034196A1; JP2018514759A; US10379027B2; WO2016151248A1; KR102479862B1; EP3274689B1; US20180080760A1; CN107532989B; JP6727229B2; FR3034196B1; CN107532989A; KR20180011762A

Abstract

The invention is a method for identifying a particle contained in a sample, including illuminating the sample using a light source, the latter producing an incident light wave propagating toward the sample, then acquiring, using a matrix-array photodetector, an image of the sample, the sample being placed between said light source and the photodetector in such a way that the matrix-array photodetector is exposed to a light wave that is the result of interference between the incident light wave and a diffraction wave produced by each particle. The method is characterised in that it includes applying a numerical reconstruction algorithm to the image acquired by the photodetector, in order to estimate a characteristic quantity of the light wave reaching the detector, at a plurality of distances from the detector. The variation in the characteristic quantity as a function of distance allows the particle to be identified.

Description

Particle analysis method

Description

TECHNICAL AREA

The invention is in the field of counting and identifying particles present in a liquid and in particular a body fluid, for example blood.

PRIOR ART

Body fluids, particularly blood, may comprise particles, for example cells, the number and type of which is useful to know. In the blood, for example, type and blood count, or blood count, analyzes are commonly performed in laboratory analysis. This type of analysis makes it possible to determine the number and to identify the main constituents of the blood, in particular cells of the red blood cell line, white blood cells, or platelets. These examinations are commonly performed, using powerful automata, but we are looking for simpler, less expensive methods, to obtain comparable performance.

One of the areas of research is the use of simple optical methods, such as lensless imaging. The observation of biological particles by lensless imaging has been developing since the late 2000s. This technique involves inserting a sample between a light source and a matrix photodetector, or image sensor. The image collected on the photodetector is formed by interference between the incident wave produced by the light source and the wave diffracted by the particles making up the sample. This image is frequently referred to as "hologram". Thus, for each particle, a diffraction pattern, or diffraction pattern, of its own can be recorded at the sensor. Applied to biological samples, this technique has been described in WO2008090330. It is then possible to perform a simple analysis of each particle, by comparing the diffraction pattern that it generates with previously established diffraction patterns, corresponding to known particles. But this method can find limits when the concentration of particles increases. Indeed, the counting and the identification of particles on the sole basis of the diffraction patterns detected by the image sensor shows a certain limit when the concentration of particles in the sample increases. In particular, when the sample is blood, and the particles are red blood cells, in excess of 100,000 particles per μΙ, their count is no longer reliable, as reported in the publication Seo Sungjyu, "High -throughput lensfree blood analysis on a chip ", Anal Chem, 2010 June 1. It is possible to apply so-called digital holographic reconstruction mathematical treatments in order to establish an image, called complex image, of each particle present in the sample . This method consists in retro-propagating the light wave in the object plane, in which the particles are located, the latter being placed at a known distance from the imager. The publication cited above shows that such a holographic reconstruction makes it possible to count red blood cells present in high concentration in a sample. This publication shows that, on the reconstructed complex image, the white blood cells, having undergone preliminary labeling, have a signature different from that of the red blood cells.

US2014 / 0327944 also describes a method for classifying particles, for example blood particles, on the basis of holograms, by comparing holograms acquired by an image sensor with a library of simulated holograms. But this process faces the same limits, namely a delicate implementation when the density of the particles is high.

Methods for reconstructing a complex image of cells, in this case spermatozoa, are also described in US2012 / 0148141, or WO2014 / 012031. But these methods make it possible to estimate properties of said cells, as well as their trajectory, from the reconstructed complex image. The same is true of document US2009 / 0290156, which describes a classification of particles on the basis of a complex image of said particles, as well as a tracking of the trajectory of said particles. However, the use of a complex image of a sample may be insufficient to identify a particle. A method of identifying particles, and in particular blood cells, which can be applied to samples in which the concentration of particles is high is sought. The method must also have an extended field of view, and must be simple to implement, in particular by avoiding prior particle marking. Furthermore, the method must make it possible to reliably discriminate particles that may be in a body fluid, in particular cells of the red blood cell, white blood cell and platelet lines. In addition, a method is sought that does not require precise knowledge of the distance between the particles and the photodetector.

SUMMARY OF THE INVENTION

The invention addresses this problem by providing a method for identifying a particle present in a sample, for example a sample of a biological fluid, such as blood, the method comprising the following steps:

illuminating said sample with a light source, the light source producing an incident light wave propagating toward the sample along an axis of propagation;

- Acquisition, using a matrix photodetector, of an image of the sample, the sample being disposed between said light source and said photodetector, such that the matrix photodetector is exposed to a light wave comprising interference between the incident light wave and a diffraction wave produced by each particle;

the method being characterized in that it also comprises the following steps:

determining a position of said particle in a plane parallel to a plane along which the matrix photodetector extends;

applying a digital reconstruction algorithm to said acquired image, so as to estimate at least one characteristic magnitude of said light wave to which the matrix photodetector is exposed, at a plurality of reconstruction distances thereof;

determining a profile, representing an evolution of said characteristic quantity as a function of said reconstruction distance, along an axis parallel to said axis of propagation and passing through said position;

- identification of the particle according to said profile.

By applying a digital reconstruction algorithm is meant the application of a propagation operator to an image, generally in the form of a convolution product.

The characteristic quantity can be obtained by estimating, at each reconstruction distance, a complex expression of the light wave to which the matrix photodetector is exposed. The characteristic quantity can be determined from the module of said complex expression, in which case it is representative of the amplitude of said light wave to which the detector is exposed. The characteristic quantity can be determined from the argument of said complex expression, in which case it is representative of the phase of said light wave to which the matrix photodetector is exposed.

According to one embodiment, the method comprises:

the determination of a complex image, referred to as a complex reference image, by applying a digital reconstruction algorithm to the image acquired by the matrix photodetector;

from said complex reference image, the estimation of at least one characteristic quantity of the light wave to which the matrix photodetector is exposed at a plurality of reconstruction distances of the latter.

The process can then comprise:

applying a propagation operator to the complex reference image, so as to calculate so-called secondary complex images, according to a plurality of distances from the reconstruction plane or the plane along which the matrix photodetector extends; determining a characteristic quantity at each of said distances from each secondary complex image.

The complex reference image may be a complex image formed in a reconstruction plane, distant from the plane of the sample. It can also be a complex image formed in the detection plane.

The identification can be developed by comparing the evolution of said characteristic quantity with typical profiles determined during a learning phase.

The position of each particle, in a plane parallel to the plane of the matrix photodetector, can be determined using the image acquired by the photodetector or by means of the complex expression of the light wave to which is exposed the photodetector.

The light source is preferably a spatially coherent source, and for example a light-emitting diode, in which case a spatial filter is preferably disposed between the light source and the sample. The light source may be temporally coherent, for example being a laser diode.

The matrix photodetector comprises an array of pixels capable of collecting the wave to which the photodetector is exposed. The distance between the pixels and the sample may vary between 50 μιη and 2 cm, and preferably between 100 μιη and 5 mm. Preferably, the sample is not disposed in direct contact with the pixels of the photodetector. Preferably, no magnification optics are disposed between the sample and the matrix photodetector.

The sample may in particular comprise blood cells. In this case, the particles can be identified among cell lines of white blood cells, red blood cells or platelets.

Another object of the invention is a device for identifying a particle, said particle being contained in a sample, the device comprising:

a light source arranged to produce an incident light wave, along an axis of propagation, towards said sample;

a support for holding the sample between said light source and a matrix photodetector;

the matrix photodetector, arranged to acquire an image of the sample, the matrix photodetector being able to be exposed to a light wave, resulting from the interference between said incident light wave and a diffraction wave formed by said particle;

characterized in that the device comprises a processor, of the electronic computer or microprocessor type, configured to implement the following operations:

determining a position of said particle in a plane parallel to the plane of the matrix photodetector;

determining a profile, representing the evolution of said characteristic quantity as a function of said reconstruction distance, along an axis parallel to said axis of propagation and passing through said position;

identifying the particle according to said profile.

Preferably, the device has no magnification optics between the matrix photodetector and the analyzed sample.

The processor may comprise or be connected to a programmable memory, comprising a sequence of instructions for implementing the previously described steps. It can in particular be able to: to determine, at each reconstruction distance, the complex expression of the optical radiation to which the detector is exposed,

estimating said characteristic quantity, at each reconstruction distance, by determining the module or the argument of said complex amplitude.

FIGURES

Figure 1 shows the device according to one embodiment of the invention.

FIG. 2A represents an image acquired by the matrix photodetector.

FIG. 2B represents the profile of a characteristic quantity, called complementary amplitude, of the light wave to which the photodetector is exposed, as a function of the distance with respect to the photodetector, according to a first example, for different types of particles.

FIG. 3A represents the profile of a characteristic quantity, called complementary amplitude, of the light wave to which the photodetector is exposed, as a function of the distance relative to the photodetector, for different white blood cells, according to this first example.

FIG. 3B represents the profile of a characteristic quantity, called complementary amplitude, of the light wave to which the photodetector is exposed, as a function of the distance with respect to the photodetector, for different red blood cells, according to this first example.

FIG. 4A represents a region of interest of the image acquired by the photodetector, centered on an aggregate of platelets, according to this first example.

FIG. 4B represents the profile of a characteristic quantity, called complementary amplitude, of the light wave to which the photodetector is exposed, for different plates forming part of the aggregate represented in FIG. 4A.

FIG. 5A represents the profile of a characteristic quantity, called complementary amplitude, of the light wave to which the photodetector is exposed, as a function of the distance with respect to the photodetector, for different types of particles, according to a second example. FIG. 5B represents the profile of the phase of the light wave to which the photodetector is exposed, as a function of the distance z with respect to the photodetector, for different types of particles, according to a second example.

FIG. 6A represents the profile of a characteristic quantity, called complementary amplitude, of the light wave to which the photodetector is exposed, as a function of the distance with respect to the photodetector, for different types of particles, according to a third example.

FIG. 6B represents the profile of the phase of the light wave to which the photodetector is exposed, as a function of the distance with respect to the photodetector, for different types of particles, according to a third example. FIGS. 7A, 7B and 7C respectively represent the profile of a composite quantity characteristic of the light wave to which the photodetector is exposed, as a function of distance, according to the first, second and third examples.

FIGS. 8A, 8B, 8C and 8D represent respectively:

a method for calculating a complex image, referred to as a complex reference image, of a sample in a reconstruction plane;

a hologram acquired by the matrix photodetector;

a representation of an image, called complex reference image, reconstructed after several iterations of the method shown in FIG. 8A.

a profile obtained on the basis of secondary complex images formed from the complex reference image.

FIG. 9A is a hologram acquired by an image sensor, the sample comprising red blood cells dispersed in an aqueous solution. FIGS. 9B and 9C respectively represent the module and the phase of a complex image, referred to as a reference image, this complex image being formed in a reconstruction plane. FIGS. 9D and 9E are profiles respectively representing an evolution of the modulus and phase of the light wave to which the image sensor is exposed, along an axis of propagation passing through a red cell.

DESCRIPTION OF PARTICULAR EMBODIMENTS

FIG. 1 represents an exemplary device that is the subject of the invention. A light source 11 is able to produce a light wave 12, referred to as an incident light wave, in the direction of a sample 14, along an axis of propagation Z. The sample 14 comprises a medium 10, for example a biological fluid, comprising particles 1, 2, 3, 4, 5, .., 9, which it is desired to identify among predetermined types of particles. A particle can be a cell. In particular, when the medium is blood, or a solution comprising blood, a particle may be a red blood cell, a white blood cell or a wafer.

A particle may also be an organic or inorganic microbead, for example a metal microbead, a polymer or glass microbead, this type of microbead being commonly used in the production of biological protocols. A particle may also be a droplet, for example a lipid droplet, immersed in the medium 10. It may also be a microorganism, for example a bacterium or a yeast, or an exosome. In general, a particle has a size advantageously less than 1 mm, or even less than 500 μιη, and preferably a size between 0.5 μιη and 500 μιη. Thus, the term particle refers to both endogenous particles, initially present in the sample examined, and exogenous particles, added to this sample prior to analysis.

The medium is most frequently a liquid medium, and especially a body fluid, but it can also act agar, or air, or the dry residue of a liquid.

The method which is the subject of the invention makes it possible to identify each particle observed. By identification is meant the classification of the particle into a predetermined particle class. It can be a determination of a nature of a particle among predetermined natures, or a determination of a size of a particle among predetermined natures. The distance Δ between the light source and the sample is preferably greater than 1 cm. It is preferably between 2 and 30 cm. Preferably, the light source, seen by the sample, is considered as point. This means that its diameter (or diagonal) is preferably less than one-tenth, better one-hundredth of the distance between the sample and the light source. Thus, the light arrives at the sample in the form of plane waves, or can be considered as such.

The light source 11 may be punctual, or be associated with a diaphragm, or spatial filter, not shown in Figure 1, so as to appear punctual. The opening of the diaphragm is typically between 5 μιη and 1mm, preferably between 50 μιη and 500 μιη.

The diaphragm may be replaced by an optical fiber, a first end of which is placed facing a light source, and a second end of which is placed facing the sample. In this case, said second end can be likened to a point light source 11.

The sample 14 is delimited by an enclosure having a bottom 15 and a cover 13. The side walls of the enclosure are not shown. In the example, the chamber is a fluid chamber Neubauer C-chip. The distance between the bottom 15 and the cover 13 is 100 μιη. In general, the thickness of the enclosure, along the axis of propagation Z, is less than a few cm, for example less than 1 cm, or even less than 1 mm, for example between 50 μιη and 500 μιη .

The light source 11 may be temporally coherent but this is not necessary. In this first example, the light source is a laser diode emitting at the wavelength of 450 nm. It is located at a distance of 15 cm from the sample. The sample 14 is disposed between the light source 11 and a matrix photodetector, or image sensor, 16. The latter preferably extends parallel to, or substantially parallel to, the bottom 15 of the enclosure delimiting the sample.

The term substantially parallel means that the two elements may not be strictly parallel, an angular tolerance of a few degrees, less than 10 ° being allowed.

Preferably, the light source is of small spectral width, for example less than 100 nm, or even 20 nm and even more preferably less than 5 nm. The term spectral width refers to the half-height width of the emission peak of the light source. The photodetector 16 may be a matrix photodetector comprising a matrix of pixels, of the CCD type or a CMOS. CMOS are the preferred photodetectors because the size of the pixels is smaller, which makes it possible to acquire images whose spatial resolution is more favorable. In this example, the detector is a 12-bit APTINA sensor, reference MT9P031. It is a CMOS RGB sensor, whose inter-pixel pitch is 2.2 μιη. The effective area of the photodetector is 5.7 x 4.3 mm ² .

The photodetector extends along a detection plane P, preferably perpendicular to the propagation axis Z of the incident light wave 12.

Preferably, the photodetector comprises a matrix of pixels, above which is disposed a transparent protection window. The distance between the pixel matrix and the protection window is generally between a few tens of μιη at 150 or 200 μιη. Photodetectors whose inter pixel pitch is less than 3 μιιι are preferred, in order to improve the spatial resolution of the image.

The distance d between the particles 1,2, ... 9 and the pixel matrix of the photodetector 16 is, in this example, equal to 1.5 mm. But it can fluctuate depending on the thickness of the fluidic chamber used. In general, and whatever the embodiment, the distance d between a particle and the pixels of the photodetector is preferably between 50 μιη and 2 cm, preferably between 100 μιη and 2 mm.

Note the absence of magnification optics between the matrix photodetector 16 and the sample 14. This does not prevent the possible presence of microlenses focusing at each pixel of the photodetector 16. In this first example, the sample is plasma enriched in white blood cells, obtained according to a usual protocol, after sedimentation of the red blood cells in the presence of Dextran (Sigma Aldrich reference D4876) 6% in Alsever solution, then recovery of the rich plasma white blood cells and in platelets. The plasma obtained is then diluted in a PBS buffer at physiological pH (acronym for Phosphate Buffer Saline - Saline Phosphate Buffer). The depletion of red blood cells is not complete and the enriched plasma obtained contains residual red blood cells.

Particles can be classified into several types of particles, including red blood cells, white blood cells or platelets. Preferably, the particles do not undergo any prior marking.

FIG. 2A represents an image obtained by the photodetector 16. This image represents a global diffraction figure, in which elementary diffraction figures can be distinguished, each elementary diffraction figure being respectively associated with the particles. Each elemental diffraction pattern comprises a disk-shaped central area around which concentric, alternately dark and clear rings extend. Such an elementary figure allows the selection of a particle to be identified, as well as the determination of the coordinates (x, y) of said particle in the detection plane P, said radial coordinates. These coordinates are for example the center of the elementary diffraction pattern corresponding to said particle. Each elementary diffraction pattern is formed by the interference between the incident light wave 12 produced by the source 11, upstream of the sample, and a wave resulting from the diffraction of the incident wave by a particle. Thus, the photodetector 16 is exposed to a light wave 22 formed by the superposition:

the light wave 12 emitted by the source 11, upstream of the sample 14,

the light wave diffracted by each of the particles or other diffracting elements present in the sample 14.

A processor 20, for example a microprocessor, receives the images of the matrix photodetector 16, and performs a reconstruction of characteristic quantities of the light wave 22 to which the matrix photodetector is exposed, along the axis of propagation Z. The microprocessor 20 is connected to a memory 23 able to store instructions for implementing the calculation steps described in this application. It can be connected to a screen 25. The In particular, reconstruction is performed between the matrix photodetector and the observed sample.

The processor 20 may be able to execute a sequence of instructions stored in a memory, for the implementation of the steps of the identification method. The processor may be a microprocessor, or any other electronic computer capable of processing the images provided by the matrix photodetector, to perform one or more steps described in this description.

The image / acquired by the matrix photodetector shown in FIG. 2A represents the spatial distribution of the intensity I (x, y) of the light wave 22, where x and y designate the coordinates in the plane P of the photodetector. .

According to the well-known principles of digital holographic reconstruction, described in Yle et al., "Digital in-line holography of biological specimens", Proc. Of SPIE Vol.6311 (2006), by carrying out a product of convolution between the intensity I (x, y) measured by the photodetector, and a propagation operator h (x, y, z), it is possible to reconstruct a complex expression U (x, y, z) of the light wave 22 at any coordinate point (x, y, z) of the space, and in particular in a plane at a distance | z | of the photodetector.

The propagation operator h (x, y, z) has the function of describing the propagation of light between the photodetector 16 and a coordinate point (x, y, z). It is then possible to determine the amplitude u (x, y, z) and the phase ç x, y, z) of this light wave at this distance I z I, called the reconstruction distance, with:

u (x, y, z) = abs [U (x, y, z)],

- φ χ, y, z) = arg [U (x, y, z)]

The operators abs and arg denote respectively the module and the argument.

The application of the propagation operator makes it possible in particular to estimate the complex expression at a distance | z | of the photodetector, upstream of the latter. The complex value of the light wave 22 is then reconstructed before the latter reaches the detector. This is called back propagation. By assigning the coordinate z = 0 to the detection plane P, this backpropagation is implemented by applying a propagation operator h (x, y, - | z |). The terms upstream and downstream are to be understood according to the direction of propagation of the incident wave 12. If I (x, y) = I (x, y, z = 0) corresponds to the intensity of the signal measured by the photodetector, a relation between the measured intensity I (x, y) and the complex expression of the The light wave U (x, y) at the detection plane P is given by: I (x, y) = \ U (x, y) \ ² .

The complex expression of the light wave (22) at a coordinate (x, y, z) is given by U (x, y, z) = J / (x, y) * h (x, y, z) ), the symbol * denoting a convolution product. with: z <0 in the half-space delimited by the detection plane P, comprising the sample 14

z> 0 in the half-space delimited by the detection plane P and not including the sample 14.

In the half-space delimited by the detection plane P and comprising the sample 14, the complex expression of the light wave can also be written:

U (x, y, z) = jl {x, y) * h (x, y, - \ z \)

Preferably a mathematical preprocessing is previously applied to the measured intensity I (x, y) before the holographic reconstruction. This improves the quality of results, including reducing artifacts when applying the propagation operator.

Thus, we determine an intensity I (x, y), called normalized intensity, such that

7 (x, y) = (I (x, y) - Average (I)) / Average (I) with

I (x, y) = intensity measured by the photodetector at the (x, y) coordinate,

Average (/) = mean of the intensity measured in a region of interest of the image /, including said coordinate (x, y). This region of interest may correspond to the entire image formed by the photodetector. This pre-treatment is similar to a normalization of the intensity measured by the intensity of the incident light wave (12), the latter being estimated by the operator Average (/).

Then the complex expression of the wave (22) is determined from the normalized intensity I (x, y) according to the equation U (x, y, z) = J (x, y) * hx, y, z) as previously explained. Digital reconstruction can be based on the Fresnel diffraction model.

In this example, the propagation operator is the Fresnel-Helmholtz function, such that:

,. x

h (x, y, z) = -e ^TT exp (jn ^ -).

where is the wavelength. Thus, U (x, y, z) = -j- e ^I27r (Yi (x ', y') ex (jn ^ ^{~ x} '^ ^ ^{y ~ y} ' ^) dx'dy '

or

x 'and y' denote the coordinates in the plane of the photodetector, x and y denote the coordinates in the reconstruction plane, the latter being situated at a distance | z | of the photodetector, z denotes the coordinate of the image reconstructed along the propagation axis Z of the incident light wave (12).

From the values of the complex expression U (x, y, z), it is possible to extract characteristic quantities of the light wave 22 resulting from the diffraction, by the particles (1,2.9), of the incident light wave 12 emitted by the source 11. As previously mentioned, it is possible to evaluate the amplitude u (x, y, z) or the phase cp x, y, z), but it is also possible to evaluate any function of amplitude or phase.

One can, for example, evaluate a characteristic quantity, called complementary amplitude, û (x, y, z) such that:

û (x, y, z) = abs (l - U (x, y, z))

From each reconstructed complex expression U (x, y, z), it is possible to constitute: an image u _z of the amplitude of the wave 22, along a plane parallel to the detector plane, at a distance | z | of the latter, with u _z (x, y) = abs [U (x, y, z)], an image φ _ζ of the phase of the wave 22, along a plane parallel to the plane of the detector, at a distance | z | of the latter, with φ _ζ (x,) = arg [U (x, y, z)], an image i½ of the complementary amplitude, as previously defined, of the wave 22, in a plane parallel to the detector plane, at a distance | z | of the latter, with i½ (x, y, z) = abs [1- U (x, y, z)].

In this first example, an image u _z of the amplitude complementary to a plurality of coordinates zi ... z _M is reconstructed along the axis of propagation Z, M being here equal to 21,

from each image ΊΣ ^, with 1 <m <M, the value û (x _n , y _n , z _m ) is extracted,, y _n ) representing the coordinates of the particle n in a plane parallel to the plane n of the photodetector 16,

different values of û (x _n , y _n , z) are obtained, with z _m <z <z _m = ₊ i by interpolation between two quantities û (x _n , y _n , z _m ) and û (x _n , yn, z _m + i) previously determined.

The coordinates (x _n , y _n ), in a plane parallel to the plane of the photodetector 16, of each particle n examined, are determined either by means of the acquired image I (x, y) or from a image û _z at a given reconstruction height z.

FIG. 2B represents, for different types of particles, the evolution of the complementary amplitude û (x _n , y _n , z), as previously defined, as a function of the distance | z | of reconstruction, for 9 different particles:

particles 1 to 4: white blood cells designated by the acronym WBC,

particles 5 and 6: red blood cells, designated by the acronym BC,

particles 7 to 9: platelets, designated by the letter PLT.

The distance | z | reconstruction varies between 1500 μιη.

Parallel to these operations, each particle (1, ..., 9) was observed under a microscope, observation under the microscope serving as a reference measurement, allowing a certain identification.

It is observed that for the sample studied in this example:

for particles 1 to 4, which are WBC white blood cells, the curve û (z) representing the evolution of the complementary amplitude as a function of the reconstruction distance has a minimum less than a threshold of amplitude θthreshoid, followed by a rise towards the baseline BL, this ascent presenting marked oscillations;

for particles 5 and 6, corresponding to RBC red blood cells, the curve û (z) has a minimum between baseline BL and amplitude threshold θthreshoid, followed by a monotonous rise towards baseline BL;

for particles 7 to 9, corresponding to platelets PLT, the curve û (z) follows the baseline BL and remains confined between two values BL ± ε. Thus, for each particle n detected, whose position along a plane parallel to the plane of the detector is (x _n , y _n ), it is possible to establish a profile û (x _n , y _n , z) representing the evolution amplitude complementary to a plurality of reconstruction heights z, and use this profile to perform an identification of the particle between a red blood cell, a white blood cell and a wafer.

This profile can in particular be compared to a library of profiles made, during a learning phase, on known particles. In other words, the profile û (z) representing the evolution of the complementary amplitude, along the axis of propagation Z (z coordinate axis), constitutes a signature of the type of the observed particle. Contrary to the prior art, a complex image of a particle is not formed by carrying out a holographic reconstruction at a predetermined distance from the sample, but a characteristic of the wave 22, resulting from the diffraction of the sample, is reconstructed. a particle with the incident wave 12, along the propagation direction of the incident wave, at a plurality of distances from the photodetector. The information obtained is richer and allows a clear classification between different types of particles.

FIGS. 3A and 3B show complementary amplitude profiles û (z) of the wave 22 to which the detector is exposed, obtained respectively for 50 WBC white cells and 240 RBC red cells. Sufficient repeatability of the profiles is observed to allow a robust classification of the particles on the basis of this profile. These profiles were obtained under experimental conditions similar to those of the preceding example.

The sample used to make the measurements shown in Figure 3A is an enriched plasma similar to the sample described in connection with Figures 2A and 2B.

The sample used to carry out the measurements shown in FIG. 3B comprises whole blood diluted in a phosphate phosphate buffer previously mentioned, dilution factor 1/400. FIG. 4B represents complementary amplitude profiles û (z), along the Z axis, obtained for 4 platelets 101, 102, 103, 104, on a sample of enriched plasma type as previously described. Microscopic observation showed that platelets 101, 102, 103 and 104 were aggregated.

FIG. 4A represents a region of interest of the image I acquired by the photodetector 16. It makes it possible to identify the coordinates (xioi, yioi), (xio2, yio2), (xio3, yio3), (xio4, yio4) of each plate of the aggregate. These profiles were obtained under experimental conditions similar to those of the first example. It is observed that the profile û (z) is similar whether the platelets are aggregated or not, and remains confined around a baseline BL, in a gap BL ± ε. Thus, according to the identification method that is the subject of the invention, the platelets are correctly identified whether they are aggregated or not.

In a second example, the light source 11 is a white light-emitting diode coupled to an Omega optical filter 485-DF-22, centered on the wavelength λ = 485ηιη and with a width at mid-height equal to 22 nm. The distance Δ between the light source and the detector is 8 cm. The sample is an enriched plasma as previously described. According to this example, a reconstruction of the complex expression U (x, y, z) of the wave 22 has been carried out at a plurality of distances z of the detector, and then at coordinates (x, y) of different particles, the complementary amplitude and the phase of the radiation. We then established profiles û (z) and φ (ζ) of the complementary amplitude and of the phase as a function of z. As in the previous examples, the nature of the observed particles was confirmed by microscopic observation.

FIGS. 5A and 5B respectively represent the complementary amplitude û (x, y, z) and the phase φ (χ, γ, ζ) as a function of the distance z for different particles. In FIG. 5A, it can be seen that:

for the particles corresponding to WBC white blood cells, the curve û (z) representing the evolution of the complementary amplitude as a function of the reconstruction distance has a marked minimum less than a threshold of amplitude θthreshoid, followed by a ascent towards the base line BL, this ascent presenting marked oscillations;

for the particles corresponding to red blood cells BC, the curve û (z) has a minimum between the base line BL and the amplitude threshold θthreshoid, then the curve follows a monotonous rise towards the baseline BL;

for the particles corresponding to platelets PLT, the curve û (z) follows the baseline BL and remains confined between two values BL ± ε.

Thus, for each particle n detected, whose position along a plane parallel to the plane of the detector is (x _n , y _n ), it is possible to establish a profile û (x _n , y _n , z) representing the evolution of the complementary amplitude û, as previously defined, of the wave 22 to which is exposed the detector, at a plurality of reconstruction heights z, and using this pattern to perform a classification of the particle between an RBC RBC, a WBC white blood cell and a PLT wafer.

Classification is therefore possible with a light source different from a laser source. In FIG. 5B, it can be seen that:

for the particles corresponding to WBC white corpuscles, the curve φ (ζ) representing the evolution of the phase φ as a function of the reconstruction distance z has a maximum greater than a first phase threshold c threshoid ¹ , then a lower minimum at a second threshoid phase threshold ² , followed by a rise towards the baseline BL;

for the RBC RBC particles, the φ (ζ) curve has a maximum greater than said first threshoid phase threshold ¹ and then a minimum greater than said second threshoid phase threshold ² , then the curve follows a monotonous rise towards the baseline BL;

for the particles corresponding to platelets PLT, the curve φ (ζ) representing the evolution of the phase φ as a function of the reconstruction distance remains confined between the two values c threshoid ¹ and c threshoid ² . The measured values remain between said first and second phase thresholds.

Thus, for each detected particle n whose position in a plane parallel to the plane of the detector is (x _n , y _n ), it is possible to establish a profile φ (χ _η , γ _η , ζ) representing the evolution of the radiation phase to which the detector is exposed, at a plurality of reconstruction heights z, and to use this profile to perform a classification of the particle between a red blood cell, a white blood cell and a wafer.

According to a third example, the light source 11 is a white light-emitting diode coupled to an Omega optical 610-DF-20 filter, centered on the wavelength λ = 610ηιη and having a half-width of 20 nm, placed at a distance of distance Δ, equal to 8 cm, of the sample. The procedure and the sample are similar to the previous example.

FIGS. 6A and 6B respectively represent the complementary amplitude û (x, y, z) and the phase φ (χ, γ, ζ) as a function of the distance z for different particles. In FIG. 6A, it can be seen that:

for the particles corresponding to WBC white blood cells, the curve û (z) representing the evolution of the amplitude u as a function of the reconstruction distance z has a marked minimum less than a threshold of complementary amplitude θthreshoid, followed by a rise towards the baseline BL, this rise having marked oscillations;

for particles corresponding to RBC red blood cells, the curve û (z) has a minimum between the base line BL and the amplitude threshold θthreshoid, then the curve follows a monotonous rise towards the baseline BL;

for the particles corresponding to platelets PLT, the curve û (z) follows the baseline BL and remains confined between two values BL ± ε

Thus, for each particle n detected, whose position along a plane parallel to the plane of the detector is (x _n , y _n ), it is possible to establish a profile û (x _n , y _n , z) representing the evolution of the complementary amplitude, as defined above, of the radiation to which the detector is exposed, at a plurality of reconstruction heights z, and to use this profile to perform a classification of the particle between an RBC red cell, a white blood cell WBC and a PLT wafer.

In FIG. 6B, it can be seen that:

for the particles corresponding to platelets PLT, the curve φ (ζ) representing the evolution of the phase φ as a function of the reconstruction distance z remains confined between two values c threshoid ¹ , threshoid ² -The measured values remain between said first and second phase thresholds.

The two preceding examples show that a complementary amplitude profile û (z) or in phase φ (ζ) make it possible to characterize the particles.

It is also possible to constitute a composite optical parameter, denoted k, combining the complementary amplitude and the phase, in the form of a ratio, for example FIGS. 7A, 7B and 7C represent the evolution of said composite optical parameter along the propagation axis Z, by implementing the respective configurations of the first example (laser source at 405 nm), of the second example (light source white LED type combined with a filter centered on λ = 485nm) and the third example (white LED light source combined with a filter centered on λ = 610nm). In each configuration, the sample observed is an enriched plasma as previously described.

On each curve, the analyzed particles are three white blood cells (WBC) and one red blood cell (BC). It is observed that, irrespective of the source, the fluctuations of the profile k z) are more important for white blood cells than for red blood cells.

In particular, it is possible to determine a first composite threshold kthreshoid ¹ and a second composite threshold kthreshoid ² , such that when the profile kz) remains lower than the first composite threshold kthreshoid ¹ and greater than the second composite threshold kthreshoid ² , the particle analyzed is a red blood cell. When the profile crosses one of these thresholds, the examined particle is identified as a white blood cell.

The application of a digital propagation operator h to an image / acquired, or hologram, by a matrix photodetector 16 may have certain limits, because the acquired image does not include any information relating to the phase. Also, prior to the profiling, it is preferable to have information relating to the phase of the light wave 22 to which the photodetector 16 is exposed. This information relating to the phase can be obtained by reconstructing an image. complex U _z of the sample 14, according to methods described in the prior art, so as to obtain an estimate of the amplitude and the phase of the light wave 22 at the plane P of the matrix photodetector 16 or in a reconstruction plan P _z located at a distance \ z \ of the latter. The inventors have developed a method based on the calculation of a complex reference image, described in connection with FIG. 8A. This process comprises the following steps:

Acquisition of an image / sample 14 by the matrix photodetector 16, this image forming the hologram (step 100).

- Calculating a complex image, called a reference, U _ref to the sample 14 in a _z reconstruction plane P or in the detection plane P, this reference complex image having phase and amplitude information of the light wave 22 to which the matrix photodetector 16 is exposed; this step is performed by applying the propagation operator h, previously described, to the acquired image / (steps 110 to 170). This complex image is referred to as a reference image because it serves as a basis for the formation of the profile on the basis of which the particle is characterized.

Selection of a radial position (x, y) of a particle in the detection plane or in a plane parallel thereto (step 180), either by using the complex reference image U _re f or the image / acquired by the photodetector 16.

Application of the propagation operator h to the complex reference image U _re f so as to calculate complex images U _re f _iZ , called secondary _images , along the propagation axis Z (step 185).

From each secondary complex image U _re f _iZ , estimation of a characteristic quantity of the light wave 22, at the radial position (x, y) of the previously selected particle, and at a plurality of distances from the reconstruction plane P _z (or detection plane P), then forming a profile representing an evolution of said characteristic quantity along the axis of propagation Z (step 190).

Characterization of the particle according to this profile. As previously indicated, this characterization can be performed by comparing the profile obtained with standard profiles obtained during a calibration phase, using standard samples. (step 200).

The algorithm shown in FIG. 8A is detailed below, the results obtained during certain steps being illustrated in FIGS. 8B to 8D. Steps 110 to 170 constitute a preferred way of obtaining a reference complex image, noted U _re f, this image representing a spatial distribution of the complex expression of the wave 22 in a reconstruction plane P _z . Those skilled in the art will understand that other algorithms for reconstructing such a complex image, and for example the algorithms cited in connection with the prior art, are also conceivable.

Step 100: image acquisition

During this step, the image sensor 16 acquires an image / of the sample 14, and more precisely of the light wave 22 transmitted by the latter, to which the image sensor is exposed. Such an image, or hologram, is shown in FIG. 8B. This image was made using a sample 14 comprising red blood cells bathed in a saline buffer, the sample being contained in a fluid chamber of thickness 100 μιη arranged at a distance d of 1500 μιη from a CMOS sensor, according to the previously described device. Step 110: Initialization

During this step, an initial image UQ ^{= 0} of the sample 14 is defined, starting from the image / acquired by the image sensor 16. This step is an initialization of the iterative algorithm described below. in connection with steps 120 to 180, the exponent k designating the rank of each iteration. The UQ ^{= 0} module of the initial image UQ ^{= 0} can be obtained by applying the square root operator to the acquired image / by the image sensor, in which case UQ ^{= 0} = ϊ ₀ ^' . In this example, the image is normalized by a term representative of the intensity of the light wave 12 incident on the sample 14. The latter may be, for example, the square root of an average / of the image /, in which case each pixel I (x, y) of the acquired image is divided by said average, so that UQ ^{= 0} =

l

The phase <o ^{= 0} of the initial image UQ ^{= 0} is either considered as zero in each pixel (x, y) or predetermined according to an arbitrary value. Indeed, the initial image UQ ^{= 0 is a} direct result of the image / acquired by the matrix photodetector 16. However, the latter does not include information relating to the phase of the light wave 22 transmitted by the sample 14 , the image sensor 16 being sensitive only to the intensity of this light wave. Step 120: propagation

During this step, the image UQ ^_1 obtained in the plane of the sample is propagated in a reconstruction plane P _Z , by the application of a propagation operator as previously described, so as to obtain an image complex U _Z , representative of the sample 14, in the reconstruction plane P _Z. The propagation is carried out by convolution of the image UQ ^_1 by the propagation operator h_ _z , so that:

the symbol * denoting a convolution product. The index - z represents the fact that the propagation is carried out in a direction opposite to the axis of propagation Z. It is called back propagation. During the first iteration (k = 1), U ^{k = 0} is the initial image determined during step 110. During the following iterations, U ^{k _1} is the complex image in the detection plane P set. day during the previous iteration.

The reconstruction plane P _z is a plane remote from the detection plane P, and preferably parallel to the latter. Preferably, the reconstruction plane P _z is a plane P ₁₄ along which the sample 14 extends. Indeed, an image reconstructed in this plane makes it possible to obtain a generally high spatial resolution. It may also be another plane, located a non-zero distance from the detection plane, and preferably parallel to the latter, for example a plane extending between the matrix photodetector 16 and the sample 14.

Step 130: Calculating an indicator in several pixels

During this step, a quantity e ^k (x, y) associated with each pixel of a plurality of pixels (x, y) of the complex image U ^k , and preferably in each of these pixels, is calculated. This quantity depends on the value U ^k (x, y) of the image U ^k , or of its module, at the pixel (x, y) at which it is calculated. It can also depend on a dimensional derivative of the image in this pixel, for example the module of a dimensional derivative of this image.

In this example, the magnitude s ^k (x, y) associated with each pixel is a modulus of a difference of the image U ^k , at each pixel, and the value 1. Such a quantity can be obtained according to the expression :

Step 140: establishing a noise indicator associated with the image U ^k .

In step 130, we have calculated quantities s ^k (x, y) in several pixels of the complex image U ^k . These quantities can form a vector E ^fe , whose terms are the quantities s ^k (x, y) associated with each pixel (x, y). In this step, an indicator, said noise indicator, is calculated from a vector standard E ^fe . In general, a norm is associated with an order, so that the norm || x || _p of order p of a vector x of dimension n of coordinates {x ₁ x _2i ■■■■ ½,) is such that: || x || _p = (Σ _{= 1} | xj with p> 0.

In the present case, a standard of order 1 is used, in other words p = 1. The inventors have indeed considered that the use of a standard of order 1, or of order less than or equal to 1, is particularly suitable. to such a sample, as explained below. During this step, the magnitude s ^k (x, y) calculated from the complex image U ^k , at each pixel (x, y) of the latter, is summed so as to constitute a noise indicator s ^k associated with the complex image U ^k .

Thus, s ^k = Σ _{(¾ y)} ^k (x, y) Due to the use of a standard of order 1, or of order less than or equal to 1, the value of the noise indicator s ^k decreases when the complex image U ^k is more and more representative of the sample. Indeed, during the first iterations, the value of the phase (p ^k (x, y), in each pixel (x, y) of the image U ^k is poorly estimated.The propagation of the image of the sample from the detection plane P to the reconstruction plane P _z is then accompanied by a significant reconstruction noise, as mentioned in connection with the prior art.This reconstruction noise is in the form of fluctuations appearing in the image Because of these fluctuations, a noise indicator s ^k , as previously defined, is all the greater as the contribution of the reconstruction noise to the reconstructed image is important. Many important aspects of this step are to determine, in the detection plane P, phase values <p ^k (x, y) of each pixel of the image of the image. sample U ^k , for obtaining, during a next iteration, an im reconstructed age U ^{k + 1} whose indicator s ^{k + 1} is lower than the indicator s ^k .

During the first iteration, as previously explained, only relevant information is available on the intensity of the light wave 22, but not on its phase. The first reconstructed image i / ^{= 1} in the reconstruction plane P _z is therefore affected by a significant reconstruction noise, due to the absence of relevant information as to the phase of the light wave 22 in the plane of detection P. Therefore, the indicator s ^{k = 1} is high. During the following iterations, the algorithm proceeds with a gradual adjustment of the phase (p ^k (x, y) in the detection plane P, so as to progressively minimize the indicator s ^k .

The image U ^k in the detection plane is representative of the light wave 22 in the detection plane P, both from the point of view of its intensity and of its phase. The steps 120 to 160 are aimed at establishing, iteratively, the value of the phase <p ^k (x, y) of each pixel of the image U ^k , minimizing the indicator s ^k , the latter being obtained on the image U ^k obtained by propagation of the image U ^{k _1} in the reconstruction plane P _z . The minimization algorithm may be a gradient descent algorithm or a conjugate gradient descent algorithm, the latter being described below.

Step 150: Adjust the value of the phase in the detection plane.

Step 150 aims to determine a value of the phase <p ^k (x, y) of each pixel of the complex image U ^k so as to minimize the indicator s ^{k + 1} resulting from a propagation of the image complex U ^k in the reconstruction plane P _z , during the next iteration k + 1. For this, a phase vector φ is established, each term being the phase <p ^k (x, y) of a pixel (x, y) of the complex image U ^k . The dimension of this vector is (N _P i _X , 1), where N _P i _X denotes the number of pixels considered. This vector is updated during each iteration, by the following update expression:

<Po (x, y) = <Po ^-1 (x <y) + a ^k p ^k (x, y) where:

^k is an integer, denoted by the term "step", and representing a distance;

p ^k is a vector of direction, of dimension (N _P i _X , 1), each term of which p (x, y) forms a direction of the gradient Vs ^k of the indicator s ^k .

This equation can be expressed in vector form, as follows:

<Po ⁼ <Po ¹ + " ^fe P ^fe

We can show that:

or :

V ^k is a gradient vector, of dimension (N _P i _X , 1), each term of which represents a variation of the indicator s ^k as a function of each of the degrees of freedom of the unknowns of the problem, i.e. say the terms of the vector φ ^. ;

PFE-i _{is an ith} t _r d _e leadership _had established in the previous iteration;

β ^{] ί} is scale factor applied to the direction vector p ^{fe_ 1} .

Each term Vs ^k (x, y) of the gradient vector Ve, is such that

where Im represents the imaginary part operator and r 'represents a coordinate (x, y) in the detection plane.

The scaling factor β ^{] ί} can be expressed in such a way that: The step a ^k may vary according to the iterations, for example between 0.03 during the first iterations and 0.0005 during the last iterations.

The updating equation makes it possible to obtain an adjustment of the vector φ, which results in an iterative update of the phase φ (x, y) in each pixel of the complex image UQ. This complex image UQ, in the detection plane, is then updated by these new values of the phase associated with each pixel. Note that the module of the complex image UQ is not modified, the latter being determined from the image acquired by the matrix photodetector 16, so that UQ (X, y) = UQ ^{= 0} (x, y). Step 160: Reiteration or algorithm output.

As long as a convergence criterion is not reached, step 160 consists in repeating the algorithm, by a new iteration of steps 120 to 160, on the basis of the complex image UQ updated during the step 150. The convergence criterion can be a predetermined number K of iterations, or a minimal value of the gradient Vs ^k of the indicator, or a difference considered negligible between two phase vectors <j "o ^{_ 1} , <Po consecutive. When the convergence criterion is reached, there is an estimate considered as correct of a complex image of the sample, in the detection plane P or in the reconstruction plane P _Z.

Step 170: Obtain the reference complex image.

At the end of the last iteration, the method may comprise a propagation of the complex image UQ resulting from the last iteration in the reconstruction plane P _Z , so as to obtain a complex reference image U _re f = U _Z. Alternatively, the complex reference image U _re f is the complex image UQ resulting from the last iteration in the detection plane P. When the density of the particles is high, this alternative is however less advantageous because the spatial resolution in the detection plane P is lower than in the reconstruction plane P _Z , especially when the reconstruction plane P _Z corresponds to a plane P ₁₄ along which the sample 14 extends.

FIG. 8C represents an image of the module u = ^{= 8} of each pixel of the reference complex image i / ^{= 8} obtained in a reconstruction plane P _Z after 8 iterations. The spatial resolution of this image allows a good identification of the radial coordinates (x, y) of each particle. Step 180: Selection of radial coordinates particle.

During this step, the radial coordinate is selected (x, y) of a particle from the reference image U _ref = U ^ ^{= 30,} for example from the image of his u _re module f ₌ Uz ^{= 30} or the image of its phase <p _re f = <Pz ^{= 3} ° - As previously mentioned, the radial coordinate term designates a coordinate in the detection plane or in the reconstruction plane. It is also conceivable to make this selection from the hologram / ₀ or from the complex image U _Q obtained in the detection plane following the last iteration. However, when the number of particles increases, it is preferable to carry out this selection on the image formed in the reconstruction plane, due to its better spatial resolution, especially when the reconstruction plane P _z corresponds to the plane of the sample P ₁₄ . FIG. 8C shows the selection of a particle, surrounded by a dashed outline.

Step 185: Applying a Propagation Operator

During this step 185, the complex reference image U _re f is propagated according to a plurality of reconstruction distances, using a propagation operator h as previously defined, so as to have a plurality of complex images. , say secondary, U _re f _iZ reconstructed at different distances from the detection plane P or the reconstruction plane P _z . Thus, this step comprises the determination of a plurality of complex images U _re f _iZ such that: Uref.z ⁼ Uref * hz ^{with z} min- ^z - ^z max-

The values z _min and z _max are the minimum and maximum coordinates, along the Z axis, according to which the complex reference image is propagated. Preferably, the complex images are reconstructed according to a plurality of coordinates z between the sample 14 and the image sensor 16. The complex images can be formed on either side of the sample 14. These secondary complex images are established by applying a holographic reconstruction operator h to the reference image U _re f. However, the latter is a complex image correctly describing the light wave 22 to which the image sensor is exposed, in particular at its phase, following the iterations of steps 120 to 160. Therefore, the secondary images U _re f _iZ form a good descriptor of the propagation of the light wave 22 along the axis of propagation Z.

Step 190: Form a profile During this step, from each secondary complex image U _re f _iZ , a characteristic quantity, as previously defined, of the light wave 22 is determined so as to determine a profile representing the evolution of said characteristic quantity according to the propagation axis Z. The characteristic quantity may be, for example the module or the phase, or their combination. FIG. 8D represents the evolution of the phase <p (z) of the light wave 22 along the axis of propagation Z.

Step 200: Characterization

The particle can then be characterized from the profile formed in the previous step. Preferably, there is a database of standard profiles formed during a learning phase using known standard samples. The characterization is then performed by a comparison or classification of the profile formed on the basis of the standard profiles.

This embodiment was tested on samples with red blood cells. Another example is shown in Figures 9A-9E. In these examples, the sample comprises red blood cells diluted in an aqueous solution comprising a buffer PBS (Saline Phosphate Buffer) diluted 1/400. The sample 14 was placed in a fluid chamber 15 with a thickness of 100 μm, placed at a distance of 8 cm from the light-emitting diode previously described, whose spectral band is centered on 450 nm. The sample is placed at a distance of 1.5 mm from the previously described CMOS image sensor. The opening of the spatial filter 18 is 150 μιη. Figure 9A shows the image / acquired by the image sensor. The images of the module and the phase of the reconstructed complex image U _z ^{= 8} , in the plane of the sample P ₁₀ , are respectively represented in FIGS. 9B and 9C. These images were obtained in 8 iterations.

The image U _z ^{= 8} constitutes a reference image U _re f, to which the propagation operator h previously described has been applied, so as to have a plurality of secondary complex images A _re f _Z according to FIG. propagation axis Z. In addition, on the image of the module or on the image of the phase of the reference image, a red blood cell has been identified, the latter being surrounded by dots on each of these images. The radial coordinates (x, y) of this red blood cell have been extracted. From the complex secondary images A _re f _Z , a profile u (z) representative of the module and a profile φ (ζ) representative of the phase of the light wave 22 reaching the image sensor 16 have been formed. The value each point of the profile is respectively obtained by determining the module and the phase of a secondary image to said radial coordinates. FIGS. 9D and 9E respectively represent the profile of the module and the phase of the red blood cell thus selected. The profile was determined between coordinates z _min = 1000 μιη and z _max = 2000 μιη with a step in z of 5 μιη. The reconstruction plane is located at 1380 μιη from the detection plane, which corresponds to the abscissa 76 in FIGS. 9D and 9E. The described method is not limited to blood and can be applied to other body fluids, eg urine, cerebrospinal fluid, bone marrow, etc. Moreover, the method can be applied to non-bodily fluids, in particular for the analysis of pollutants or toxins in water or other aqueous solution.

The method also applies to the detection and identification of particles placed in a non-liquid medium, for example an agar or the dry residue of a body fluid, for example a blood smear resulting in an extensive deposition of dry blood. on a blade. In the latter case, the particles are isolated from each other by dry residues or air.

Moreover, as previously indicated, the particles may be endogenous (for example blood particles) or exogenous (microbeads, droplets). The examples described above expose simple identification criteria, based on the evolution of the profile of a characteristic quantity as a function of the reconstruction distance, and comparisons using pre-established thresholds. The validity of the criteria is related to the medium in which the particles are placed, as well as to the sample preparation protocol. Other criteria may apply to particles having undergone a different preparation protocol. Thus, for a given type of sample, the identification criteria can be defined during a learning phase, carried out on standard samples, comprising known particles.

In addition, other classification methods, more complex, and more robust, can be implemented, without departing from the scope of the invention.

Claims

A method for identifying a particle (1, 2, ... 9, 101 ... 104) contained in a sample (14), the method comprising the steps of: a) illuminating said sample with a light source (11), the light source producing an incident light wave (12) propagating towards the sample (14) along an axis of propagation (Z); b) acquiring, with the aid of a matrix photodetector (16), an image of the sample, the sample being disposed between said light source and the matrix photodetector, so that the matrix photodetector is exposed a light wave (22) comprising interference between the incident light wave (12) and a collision wave produced by each particle; the method being characterized in that it also comprises the following steps: c) determining a position of said particle in a plane parallel (x, y) to a plane (P) along which the matrix photodetector (16) ); d) applying a digital reconstruction algorithm to said acquired image, so as to estimate at least one characteristic magnitude (u, ΰ, φ, k) of said light wave (22) to which the matrix photodetector (16) is exposed at a plurality of distances (| z |) of reconstruction of the latter; e) determining a profile, representing an evolution (u (z), ΰ (ζ), φ (ζ), k (z)) of said characteristic quantity as a function of said distance, along an axis parallel to said axis of propagation (Z) through said position (x, y); f) identifying the particle according to said profile.

2. Method according to claim 1, wherein said characteristic quantity is obtained by estimating, at each reconstruction distance, a complex expression (U (x, y, z)) of the light wave (22) to which is exposed the matrix photodetector.

The method of claim 2, wherein the characteristic quantity is determined from the module or argument of said complex expression (U (x, y, z)).

4. The method according to claim any one of the preceding claims, wherein the identification is made by comparing the evolution of said characteristic quantity with typical profiles determined during a learning phase.

5. Method according to any one of claims 2 to 4, wherein the position of each particle in a plane parallel to the plane of the matrix photodetector, is determined using the image acquired by the photodetector or at the using the complex expression (U (x, y, z)) of the light wave (22) to which the matrix photodetector is exposed.

6. Process according to any one of claims 1 to 5, comprising:

determining a reference complex image (i _re /), in a reconstruction plane or in the detection plane, by applying a digital reconstruction algorithm to the image acquired by the matrix photodetector (16); from said complex reference image, the estimation of at least one characteristic quantity (u, ΰ, φ, k) of the light wave (22) to which the matrix photodetector (16) is exposed, to a plurality of distances (| z |) of reconstruction of the latter.

7. Process according to claim 6, comprising:

applying a propagation operator () to the complex reference image (U _re /), so as to calculate so-called secondary complex images (f / _re / _jZ ), according to a plurality of distances of the reconstruction plane or the plane along which the matrix photodetector extends;

determining a characteristic quantity at each of said distances from each secondary complex image (f / _re / _jZ ).

The method of any of the preceding claims, wherein the light source is a spatially coherent source.

The method of any of the preceding claims, wherein the light source is a light emitting diode or a laser diode.

The method of any of the preceding claims, wherein no magnification optics are disposed between the sample and the matrix photodetector.

The method of any one of the preceding claims, wherein the sample comprises blood cells.

12. A method according to any one of the preceding claims, wherein the particles are identified among cell lines of white blood cells, red blood cells or platelets.

13. A device for identifying a particle, said particle being contained in a sample (14), the device comprising: a light source (11) arranged to produce an incident light wave (12) along an axis of propagation (Z) in the direction of said sample (14); a support for holding the sample (14) between said light source (11) and a matrix photodetector (16); the matrix photodetector (16) being arranged to acquire an image of the sample, by being exposed to a light wave (22), resulting from the interference between said incident light wave (12) and a diffraction wave formed by said particle; the device being characterized in that it comprises processor (20), and in particular a microprocessor configured to implement the following steps: - determining a position of said particle in a plane parallel (x, y) to a plane

(P) according to which extends the matrix photodetector (16); applying a digital reconstruction algorithm to said acquired image, so as to estimate at least one characteristic magnitude (u, φ, φ, / c) of said light wave (22) to which the matrix photodetector (16) is exposed, at a plurality of distances (| z |) of reconstruction of the latter; determining a profile, representing the evolution (u (z), ΰ (ζ), φ (ζ), kz)) of said characteristic quantity as a function of said reconstruction distance, along an axis parallel to said axis of propagation and passing through said position (x, y); identifying the particle according to said profile.

14. Device according to claim 13, wherein the device has no magnification optics between the matrix photodetector and the sample analyzed.

15. Device according to any one of claims 13 or 14, wherein the processor is adapted to:

determining, at each reconstruction distance, the complex expression (U (x, y, z)) of the optical radiation to which the detector is exposed; estimating said characteristic quantity, at each reconstruction distance, determining the module or the argument of said complex amplitude (U (x, y, z)).